TrxG protein Trx and the PcG proteins Pc

and E(z) are associated with recently replicated DNA sequences also bound by
PCNA. In addition, Trx, Pc, and E(z) could
be found in close proximity to PCNA in a
proximity ligation assay. Thus, although
the H3K4m3 and H3K27m3 marks appear
to be lost in S phase, the enzyme
complexes that carry out these modifications remain associated with chromatin
during its replication. These data support
a model in which the H3K4 and H3K27
methyl marks are lost during DNA replication but are re-established after replication by the TrxG and PcG histone methyltransferase complexes (Figure 1B).
It is not yet clear whether Trx, Pc, and
E(z) remain associated with chromatin
during replication or rapidly reassociate
after passage of the replication fork.
Interestingly, the PcG proteins Psc and
Pc have been shown to stably bind chromatin during DNA replication in vitro
(Francis et al., 2009). Moreover, the ability

of Psc to oligomerize has lead to

a proposal in which the oligomer can
bridge the replication fork to associate
with newly replicated chromatin (Lo
et al., 2012). Alternatively, the ability of
Trx and E(z) to bind to single-stranded
DNA, as at the replication fork, could
account for their retention at sites of
replication (Krajewski et al., 2005). Finally,
it is also possible that TrxG and PcG
proteins are passed around the elongation fork by transiently interacting with
replication proteins, as observed for
other histone-modifying enzymes (Zhu
and Reinberg, 2011). It will be important
to uncover the mechanism by which
these enzyme complexes pass replication forks and to extend these studies to
other systems to determine whether the
loss of methylated histones and retention
of modifying complexes during replication are unique to rapidly dividing cells in
Drosophila embryos or conserved among
other cell types.

REFERENCES
Corpet, A., and Almouzni, G. (2009). Trends Cell
Biol. 19, 2941.
Francis, N.J., Follmer, N.E., Simon, M.D., Aghia,
G., and Butler, J.D. (2009). Cell 137, 110122.
Krajewski, W.A., Nakamura, T., Mazo, A., and
Canaani, E. (2005). Mol. Cell. Biol. 25, 18911899.
Lo, S.M., Follmer, N.E., Lengsfeld, B.M., Madamba, E.V., Seong, S., Grau, D.J., and Francis,
N.J. (2012). Mol. Cell 46, 784796.
Mohan, M., Herz, H.M., Smith, E.R., Zhang, Y.,
Jackson, J., Washburn, M.P., Florens, L., Eissenberg, J.C., and Shilatifard, A. (2011). Mol. Cell.
Biol. 31, 43104318.
Petruk, S., Sedkov, Y., Johnston, D.M., Hodgson,
J.W., Black, K.L., Kovermann, S.K., Beck, S., Canaani, E., Brock, H.W., and Mazo, A. (2012). Cell
150, this issue, 922933.
Simon, J.A., and Kingston, R.E. (2009). Nat. Rev.
Mol. Cell Biol. 10, 697708.
Suganuma, T., and Workman, J.L. (2011). Annu.
Rev. Biochem. 80, 473499.
Zhu, B., and Reinberg, D. (2011). Cell Res. 21,
435441.

LIS1 Clamps Dynein to the Microtubule

Martina Trokter1 and Thomas Surrey1,*
1London Research Institute, Cancer Research UK, 44 Lincolns Inn Fields, London WC2A 3LY, UK
*Correspondence: thomas.surrey@cancer.org.uk
http://dx.doi.org/10.1016/j.cell.2012.08.010

Cytoplasmic dynein is a motor essential for numerous mechanical processes in eukaryotic cells.
How its activity is regulated is largely unknown. By using a combination of approaches including
single-molecule biophysics and electron microscopy, Huang et al. in this issue uncover the regulatory mechanism by which LIS1 controls the activity of cytoplasmic dynein.
Cytoplasmic dynein is a microtubule
motor that carries out the majority of
tasks depending on minus-end directed
motility in the cytoplasm of most eukaryotic cells (Allan, 2011). Several accessory
proteins modulate dyneins properties
and functions. Prominent examples are
the dynactin complex, LIS1 and NudE
(Kardon and Vale, 2009). How these
cofactors regulate dyneins cellular activities is still poorly understood. In this
issue of Cell, Huang et al. (2012) unravel
the molecular mechanism by which LIS1

regulates the motility of cytoplasmic

dynein from Saccharomyces cerevisiae.
Cytoplasmic dynein is a fascinating
enzyme. It is a large complex consisting
of two dynein heavy chains and several
smaller subunits. The smaller subunits
associate with the N-terminal part of
the heavy chains, forming the cargo
binding region. The C-terminal part of
the heavy chain forms the motor domain.
Each motor domain consists of: (1) A
hexameric AAA+ (ATPase associated
with various cellular activities) ring with

the major ATP hydrolysis site located in

the AAA1 domain (Figure 1A); (2) The
microtubule binding domain (MTBD)
located at the end of an elongated antiparallel coiled coil (15 nm) called the
stalk that protrudes from AAA4 domain
(Figure 1A); (3) The linker connecting
AAA1 with the N-terminal sequence of
the heavy chain. This linker represents
the major mobile mechanical element
responsible for force generation and
directional movement of this motor (Cho
and Vale, 2012).

Cell 150, August 31, 2012 2012 Elsevier Inc. 877

Figure 1. LIS1 Uncouples Communication between the Primary ATPase and Microtubule
Binding Domains of Cytoplasmic Dynein
(A) Schematic of a dynein heavy chain. The motor domain consists of six AAA+ domains (numbered)
forming a ring (light blue), the stalk with the microtubule binding domain (MTBD) at its tip (dark blue), and
a mechanical element, the linker (magenta), connecting AAA1 and the tail (dynein cargo-binding region).
Regular motor stepping requires coordination of (i) ATP hydrolysis in AAA1, (ii) movement of the
mechanical element, and (iii) alternation of the MTBD between strongly and weakly microtubule (MT)bound states. Note: Only one monomer of the two heavy chains is shown here.
(B) LIS1 (red) binds at the AAA3/4 interface of the dynein motor domain, uncoupling the communication
between the ATPase and MTBD. With LIS1 bound, ATP hydrolysis can occur, whereas the MTBD stays
locked in a strongly microtubule bound state. How the movements of the linker are affected by LIS1
binding, remains to be investigated in the future. Nudel (green) is a regulatory protein that interacts with
both dynein and LIS1.

As for filament-dependent motors in

general, directional force production
requires intramolecular coupling of three
processes: ATP hydrolysis, movement of
a mechanical element, and cycling of the
filament binding site through states with
high and low affinity (Figure 1A). This
allows transformation of the energy
released by ATP hydrolysis into mechanical work precisely when the motor is
attached to its filament. Therefore, motors
typically make one step per one hydrolyzed ATP. Remarkably, in dynein the
major ATP hydrolysis site and the microtubule binding site are separated by
a distance of 25 nm (Cho and Vale,
2012). Therefore, information must be
transmitted over this distance through
conformational changes to enable correct
intramolecular coupling between the
biochemical and mechanical cycles.
Such conformational communication is
not unusual for AAA+ proteins, some of
which are DNA helicases or protein folding chaperones (Erzberger and Berger,
2006). For dynein, this communication
gains a level of complexity as informa-

tion also has to be transmitted through

the coiled coil to the MTBD, which is
likely achieved by relative sliding of its
individual helices (Cho and Vale, 2012).
Huang et al. (2012) now report a mechanism by which a regulator, LIS1, modulates dynein motility by interfering selectively with the usual coupling of the
ATP hydrolysis and microtubule binding
affinity cycles.
LIS1, first identified as being involved
in lissencephaly, a severe human brain
disorder (Reiner et al., 1993), is now
recognized to be a regulator that is
required for a variety of dynein-dependent
processes. They include organelle and
mRNA transport, nuclear and centrosomal positioning, mitotic spindle orientation, and kinetochore activity. In contrast
to other known proteins interacting with
dynein, LIS1 binds to dyneins motor
domain (Allan, 2011; Kardon and Vale,
2009). By using electron microscopy and
mutational analyses, Huang et al. (2012)
show in vitro and in yeast cells that LIS1
binds the AAA3/4 interface of the dynein
motor domain (Figure 1B), in contrast to

878 Cell 150, August 31, 2012 2012 Elsevier Inc.

previous proposals. Remarkably, this

interaction uncouples the ATP hydrolysis
cycle from microtubule binding affinity
changes at dyneins MTBD, as demonstrated by a combination of biochemical
analysis and fluorescence imaging. This
uncoupling effect is somewhat similar to
a clutch in a car that can mechanically
uncouple the engine from the wheels.
With LIS1 bound, dynein is arrested in
a strongly microtubule-bound state,
although ATP hydrolysis can still go on.
The authors provide evidence that
a conserved structural element at the
AAA3/4 interface of dyneinthat is also
critical for other AAA+ protein activities
(Erzberger and Berger, 2006)is most
likely involved in mediating the conformational transmission of information
from the ATP hydrolysis site to the
MTBD. This brings us one step closer to
understanding the molecular mechanism
of intramolecular communication in the
dynein ring. LIS1 binding disrupts this
communication.
What could this type of regulation be
useful for? In budding yeast, as in several
other species, LIS1 and dynein accumulate at growing microtubule ends.
LIS1-mediated induction of a strongly
bound state might help dynein to track
microtubule ends by decreasing its
dissociation rate (Huang et al., 2012).
LIS1 usually acts in combination with
the nuclear distribution protein E (NudE)
or its paralog NudE-like (Nudel) (Kardon
and Vale, 2009). Huang et al. (2012)
show that the only known budding yeast
ortholog Nudel supports LIS1s clutch
activity, in agreement with observations
in living yeast cells (Li et al., 2005). An
attractive possibility is that binding of
dynactin or other interaction partners
could unlock the LIS1/NudE-arrested
state of dynein and trigger dynein motility
once it is attached to its cargo or at the
cell cortex. In budding yeast, dyneins
major functional role is to pull on astral
microtubules once it is anchored at the
cortex to position the nucleus with its
spindle into the bud-neck during mitosis.
When LIS1 is present at concentrations
at which it is expected to occasionally
dissociate from the motor domain,
budding yeast cytoplasmic dynein walks
along the microtubule but with reduced
speed. Similar pausing and slowing
down of motility was observed previously

also for mammalian dynein adsorbed to

microspheres (McKenney et al., 2010).
LIS1 was also reported to help mammalian dynein to work against an external
load, something that has not yet been
explored for yeast dynein. This finding
suggests that the clutch effect of LIS1
binding might itself be load-dependent.
In this context it will be important to
understand how linker movements are
affected by LIS1 binding, both in the presence and absence of load. Other open
questions concern the action of NudE
and Nudel. Although they recruit LIS1 to
mammalian dynein, NudE/Nudel have
been shown to strongly reduce the LIS1induced effects on mammalian dynein
(McKenney et al., 2010; Torisawa et al.,
2011; Yamada et al., 2008), in striking
contrast to the situation in budding yeast
(Li et al., 2005; Markus et al., 2009; Huang
et al., 2012).

In the future, it will be interesting to

investigate to what extent dyneins from
different species evolved varying regulatory control mechanisms, possibly reflecting different tasks they perform in
these species. The availability of recombinant dynein also from other organisms,
including mammals, will be crucial for
dissecting the molecular mechanism of
dyneins regulation, as the elegant work
presented by Huang et al. (2012) demonstrates.

Huang, J., Roberts, A.J., Leschziner, A.E., and

Reck-Peterson, S.L. (2012). Cell 150, this issue,
975986.
Kardon, J.R., and Vale, R.D. (2009). Nat. Rev. Mol.
Cell Biol. 10, 854865.
Li, J., Lee, W.L., and Cooper, J.A. (2005). Nat. Cell
Biol. 7, 686690.
Markus, S.M., Punch, J.J., and Lee, W.L. (2009).
Curr. Biol. 19, 196205.
McKenney, R.J., Vershinin, M., Kunwar, A., Vallee,
R.B., and Gross, S.P. (2010). Cell 141, 304314.
Reiner, O., Carrozzo, R., Shen, Y., Wehnert, M.,
Faustinella, F., Dobyns, W.B., Caskey, C.T., and
Ledbetter, D.H. (1993). Nature 364, 717721.

REFERENCES
Allan, V.J. (2011). Biochem. Soc. Trans. 39, 1169
1178.
Cho, C., and Vale, R.D. (2012). Biochim. Biophys.
Acta 1823, 182191.
Erzberger, J.P., and Berger, J.M. (2006). Annu.
Rev. Biophys. Biomol. Struct. 35, 93114.

Torisawa, T., Nakayama, A., Furuta, K., Yamada,

M., Hirotsune, S., and Toyoshima, Y.Y. (2011). J.
Biol. Chem. 286, 19591965.
Yamada, M., Toba, S., Yoshida, Y., Haratani, K.,
Mori, D., Yano, Y., Mimori-Kiyosue, Y., Nakamura,
T., Itoh, K., Fushiki, S., et al. (2008). EMBO J. 27,
24712483.

A Cornucopia of Candidates for Deafness

Morag A. Lewis1 and Karen P. Steel1,*
1Wellcome Trust Sanger Institute, Genome Campus, Hinxton, Cambridge CB10 1SA, UK
*Correspondence: kps@sanger.ac.uk
http://dx.doi.org/10.1016/j.cell.2012.08.007

Many genes involved in deafness are yet to be discovered. Here, Senthilan et al. focus on the
Drosophila Johnstons organ to uncover a wide variety of genes, including several unexpected
candidates as well as those already known to underlie deafness in mice and humans.
Gene discovery is a persistent challenge.
New genes that are found by analyzing
interesting phenotypes are often not
those that would have been predicted,
whereas genes that might be expected
to be important may prove upon creation
of a null allele to be nothing of the sort.
Deafness, the most common sensory
deficit in the human population, is a prime
example of such a problematic phenotype. Many genes are known to contribute
to deafness, but there are undoubtedly
many more that have not yet been found
(http://hereditaryhearingloss.org/). In this
issue, Senthilan and colleagues make
good use of Drosophila, which up until

now had only 24 genes associated with

sensory perception of sound, to successfully screen for many more candidates (Senthilan et al., 2012).
The organ of hearing in fruit flies is
Johnstons organ (Figure 1), an array of
chordotonal sensilla in the third antennal
segment, which has a feathery arista that
serves as the sound receiver. The sensilla
of the Johnstons organ consist of mechanosensory neurons accompanied by scolopale, cap, and ligament cells, which are
all supporting cells (reviewed in Bechstedt
and Howard, 2008). Mechanosensory and
supporting cells are specified by the
basic helix-loop-helix protein Atonal (Ato)

(Jarman et al., 1993), the ortholog of

which (Atoh1 or Math1) serves the same
purpose in specifying hair cells in mammalian inner ears (Bermingham et al.,
1999). In addition to Ato, flies and mammals also share Myosin 7a, Prestin, and
several TRP channels, and the mechanical principles behind sensing of sound
waves are very similar in flies and vertebrates (reviewed in Boekhoff-Falk, 2005).
Senthilan et al. use a specific atonal null
mutant, which lacks Johnstons organ, to
carry out microarray experiments. Several
careful approaches are taken to ensure
that the data are robust, including cluster
analyses and scatter plots comparing

Cell 150, August 31, 2012 2012 Elsevier Inc. 879